Merge commit 'f81be06cf614919d71ded29b8f595bef40123ad8'

* commit 'f81be06cf614919d71ded29b8f595bef40123ad8':
  cavs: Change type of stride parameters to ptrdiff_t

Merged-by: James Almer <jamrial@gmail.com>

6 files changed, 75 insertions, 49 deletions

diff --git a/libavcodec/cavs.c b/libavcodec/cavs.c
index 10e118e55b..549fb9eefd 100644
--- a/libavcodec/cavs.c
+++ b/libavcodec/cavs.c
@@ -256,7 +256,7 @@ void ff_cavs_load_intra_pred_chroma(AVSContext *h)
     }
 }
 
-static void intra_pred_vert(uint8_t *d, uint8_t *top, uint8_t *left, int stride)
+static void intra_pred_vert(uint8_t *d, uint8_t *top, uint8_t *left, ptrdiff_t stride)
 {
     int y;
     uint64_t a = AV_RN64(&top[1]);
@@ -264,7 +264,7 @@ static void intra_pred_vert(uint8_t *d, uint8_t *top, uint8_t *left, int stride)
         *((uint64_t *)(d + y * stride)) = a;
 }
 
-static void intra_pred_horiz(uint8_t *d, uint8_t *top, uint8_t *left, int stride)
+static void intra_pred_horiz(uint8_t *d, uint8_t *top, uint8_t *left, ptrdiff_t stride)
 {
     int y;
     uint64_t a;
@@ -274,7 +274,7 @@ static void intra_pred_horiz(uint8_t *d, uint8_t *top, uint8_t *left, int stride
     }
 }
 
-static void intra_pred_dc_128(uint8_t *d, uint8_t *top, uint8_t *left, int stride)
+static void intra_pred_dc_128(uint8_t *d, uint8_t *top, uint8_t *left, ptrdiff_t stride)
 {
     int y;
     uint64_t a = 0x8080808080808080ULL;
@@ -282,7 +282,7 @@ static void intra_pred_dc_128(uint8_t *d, uint8_t *top, uint8_t *left, int strid
         *((uint64_t *)(d + y * stride)) = a;
 }
 
-static void intra_pred_plane(uint8_t *d, uint8_t *top, uint8_t *left, int stride)
+static void intra_pred_plane(uint8_t *d, uint8_t *top, uint8_t *left, ptrdiff_t stride)
 {
     int x, y, ia;
     int ih = 0;
@@ -304,7 +304,7 @@ static void intra_pred_plane(uint8_t *d, uint8_t *top, uint8_t *left, int stride
 #define LOWPASS(ARRAY, INDEX)                                           \
     ((ARRAY[(INDEX) - 1] + 2 * ARRAY[(INDEX)] + ARRAY[(INDEX) + 1] + 2) >> 2)
 
-static void intra_pred_lp(uint8_t *d, uint8_t *top, uint8_t *left, int stride)
+static void intra_pred_lp(uint8_t *d, uint8_t *top, uint8_t *left, ptrdiff_t stride)
 {
     int x, y;
     for (y = 0; y < 8; y++)
@@ -312,7 +312,7 @@ static void intra_pred_lp(uint8_t *d, uint8_t *top, uint8_t *left, int stride)
             d[y * stride + x] = (LOWPASS(top, x + 1) + LOWPASS(left, y + 1)) >> 1;
 }
 
-static void intra_pred_down_left(uint8_t *d, uint8_t *top, uint8_t *left, int stride)
+static void intra_pred_down_left(uint8_t *d, uint8_t *top, uint8_t *left, ptrdiff_t stride)
 {
     int x, y;
     for (y = 0; y < 8; y++)
@@ -320,7 +320,7 @@ static void intra_pred_down_left(uint8_t *d, uint8_t *top, uint8_t *left, int st
             d[y * stride + x] = (LOWPASS(top, x + y + 2) + LOWPASS(left, x + y + 2)) >> 1;
 }
 
-static void intra_pred_down_right(uint8_t *d, uint8_t *top, uint8_t *left, int stride)
+static void intra_pred_down_right(uint8_t *d, uint8_t *top, uint8_t *left, ptrdiff_t stride)
 {
     int x, y;
     for (y = 0; y < 8; y++)
@@ -333,7 +333,7 @@ static void intra_pred_down_right(uint8_t *d, uint8_t *top, uint8_t *left, int s
                 d[y * stride + x] = LOWPASS(left, y - x);
 }
 
-static void intra_pred_lp_left(uint8_t *d, uint8_t *top, uint8_t *left, int stride)
+static void intra_pred_lp_left(uint8_t *d, uint8_t *top, uint8_t *left, ptrdiff_t stride)
 {
     int x, y;
     for (y = 0; y < 8; y++)
@@ -341,7 +341,7 @@ static void intra_pred_lp_left(uint8_t *d, uint8_t *top, uint8_t *left, int stri
             d[y * stride + x] = LOWPASS(left, y + 1);
 }
 
-static void intra_pred_lp_top(uint8_t *d, uint8_t *top, uint8_t *left, int stride)
+static void intra_pred_lp_top(uint8_t *d, uint8_t *top, uint8_t *left, ptrdiff_t stride)
 {
     int x, y;
     for (y = 0; y < 8; y++)
diff --git a/libavcodec/cavs.h b/libavcodec/cavs.h
index fb9df15194..6bfb22ce48 100644
--- a/libavcodec/cavs.h
+++ b/libavcodec/cavs.h
@@ -227,8 +227,8 @@ typedef struct AVSContext {
     uint8_t intern_border_y[26];
     uint8_t topleft_border_y, topleft_border_u, topleft_border_v;
 
-    void (*intra_pred_l[8])(uint8_t *d,uint8_t *top,uint8_t *left,int stride);
-    void (*intra_pred_c[7])(uint8_t *d,uint8_t *top,uint8_t *left,int stride);
+    void (*intra_pred_l[8])(uint8_t *d, uint8_t *top, uint8_t *left, ptrdiff_t stride);
+    void (*intra_pred_c[7])(uint8_t *d, uint8_t *top, uint8_t *left, ptrdiff_t stride);
     uint8_t *col_type_base;
 
     /* scaling factors for MV prediction */
diff --git a/libavcodec/cavsdec.c b/libavcodec/cavsdec.c
index 73a923df30..6f4d6aca69 100644
--- a/libavcodec/cavsdec.c
+++ b/libavcodec/cavsdec.c
@@ -545,7 +545,7 @@ static inline int dequant(AVSContext *h, int16_t *level_buf, uint8_t *run_buf,
  */
 static int decode_residual_block(AVSContext *h, GetBitContext *gb,
                                  const struct dec_2dvlc *r, int esc_golomb_order,
-                                 int qp, uint8_t *dst, int stride)
+                                 int qp, uint8_t *dst, ptrdiff_t stride)
 {
     int i, esc_code, level, mask, ret;
     unsigned int level_code, run;
diff --git a/libavcodec/cavsdsp.c b/libavcodec/cavsdsp.c
index df9490ad8f..90a67e910c 100644
--- a/libavcodec/cavsdsp.c
+++ b/libavcodec/cavsdsp.c
@@ -42,7 +42,8 @@
 #define Q1 p0_p[ 1*stride]
 #define Q2 p0_p[ 2*stride]
 
-static inline void loop_filter_l2(uint8_t *p0_p,int stride,int alpha, int beta) {
+static inline void loop_filter_l2(uint8_t *p0_p, ptrdiff_t stride, int alpha, int beta)
+{
     int p0 = P0;
     int q0 = Q0;
 
@@ -62,7 +63,8 @@ static inline void loop_filter_l2(uint8_t *p0_p,int stride,int alpha, int beta)
     }
 }
 
-static inline void loop_filter_l1(uint8_t *p0_p, int stride, int alpha, int beta, int tc) {
+static inline void loop_filter_l1(uint8_t *p0_p, ptrdiff_t stride, int alpha, int beta, int tc)
+{
     int p0 = P0;
     int q0 = Q0;
 
@@ -81,7 +83,8 @@ static inline void loop_filter_l1(uint8_t *p0_p, int stride, int alpha, int beta
     }
 }
 
-static inline void loop_filter_c2(uint8_t *p0_p,int stride,int alpha, int beta) {
+static inline void loop_filter_c2(uint8_t *p0_p, ptrdiff_t stride, int alpha, int beta)
+{
     int p0 = P0;
     int q0 = Q0;
 
@@ -99,8 +102,9 @@ static inline void loop_filter_c2(uint8_t *p0_p,int stride,int alpha, int beta)
     }
 }
 
-static inline void loop_filter_c1(uint8_t *p0_p,int stride,int alpha, int beta,
-                                  int tc) {
+static inline void loop_filter_c1(uint8_t *p0_p, ptrdiff_t stride, int alpha, int beta,
+                                  int tc)
+{
     if(abs(P0-Q0)<alpha && abs(P1-P0)<beta && abs(Q1-Q0)<beta) {
         int delta = av_clip(((Q0-P0)*3+P1-Q1+4)>>3, -tc, tc);
         P0 = av_clip_uint8(P0+delta);
@@ -115,8 +119,9 @@ static inline void loop_filter_c1(uint8_t *p0_p,int stride,int alpha, int beta,
 #undef Q1
 #undef Q2
 
-static void cavs_filter_lv_c(uint8_t *d, int stride, int alpha, int beta, int tc,
-                           int bs1, int bs2) {
+static void cavs_filter_lv_c(uint8_t *d, ptrdiff_t stride, int alpha, int beta, int tc,
+                             int bs1, int bs2)
+{
     int i;
     if(bs1==2)
         for(i=0;i<16;i++)
@@ -131,8 +136,9 @@ static void cavs_filter_lv_c(uint8_t *d, int stride, int alpha, int beta, int tc
     }
 }
 
-static void cavs_filter_lh_c(uint8_t *d, int stride, int alpha, int beta, int tc,
-                           int bs1, int bs2) {
+static void cavs_filter_lh_c(uint8_t *d, ptrdiff_t stride, int alpha, int beta, int tc,
+                             int bs1, int bs2)
+{
     int i;
     if(bs1==2)
         for(i=0;i<16;i++)
@@ -147,8 +153,9 @@ static void cavs_filter_lh_c(uint8_t *d, int stride, int alpha, int beta, int tc
     }
 }
 
-static void cavs_filter_cv_c(uint8_t *d, int stride, int alpha, int beta, int tc,
-                           int bs1, int bs2) {
+static void cavs_filter_cv_c(uint8_t *d, ptrdiff_t stride, int alpha, int beta, int tc,
+                             int bs1, int bs2)
+{
     int i;
     if(bs1==2)
         for(i=0;i<8;i++)
@@ -163,8 +170,9 @@ static void cavs_filter_cv_c(uint8_t *d, int stride, int alpha, int beta, int tc
     }
 }
 
-static void cavs_filter_ch_c(uint8_t *d, int stride, int alpha, int beta, int tc,
-                           int bs1, int bs2) {
+static void cavs_filter_ch_c(uint8_t *d, ptrdiff_t stride, int alpha, int beta, int tc,
+                             int bs1, int bs2)
+{
     int i;
     if(bs1==2)
         for(i=0;i<8;i++)
@@ -185,7 +193,8 @@ static void cavs_filter_ch_c(uint8_t *d, int stride, int alpha, int beta, int tc
  *
  ****************************************************************************/
 
-static void cavs_idct8_add_c(uint8_t *dst, int16_t *block, int stride) {
+static void cavs_idct8_add_c(uint8_t *dst, int16_t *block, ptrdiff_t stride)
+{
     int i;
     int16_t (*src)[8] = (int16_t(*)[8])block;
 
@@ -260,7 +269,8 @@ static void cavs_idct8_add_c(uint8_t *dst, int16_t *block, int stride) {
  ****************************************************************************/
 
 #define CAVS_SUBPIX(OPNAME, OP, NAME, A, B, C, D, E, F) \
-static void OPNAME ## cavs_filt8_h_ ## NAME(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
+static void OPNAME ## cavs_filt8_h_ ## NAME(uint8_t *dst, const uint8_t *src, ptrdiff_t dstStride, ptrdiff_t srcStride)\
+{                                                                       \
     const int h=8;\
     const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
     int i;\
@@ -279,7 +289,8 @@ static void OPNAME ## cavs_filt8_h_ ## NAME(uint8_t *dst, const uint8_t *src, in
     }\
 }\
 \
-static void OPNAME ## cavs_filt8_v_  ## NAME(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
+static void OPNAME ## cavs_filt8_v_  ## NAME(uint8_t *dst, const uint8_t *src, ptrdiff_t dstStride, ptrdiff_t srcStride)\
+{                                                                       \
     const int w=8;\
     const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
     int i;\
@@ -311,7 +322,8 @@ static void OPNAME ## cavs_filt8_v_  ## NAME(uint8_t *dst, const uint8_t *src, i
     }\
 }\
 \
-static void OPNAME ## cavs_filt16_v_ ## NAME(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
+static void OPNAME ## cavs_filt16_v_ ## NAME(uint8_t *dst, const uint8_t *src, ptrdiff_t dstStride, ptrdiff_t srcStride)\
+{                                                                       \
     OPNAME ## cavs_filt8_v_ ## NAME(dst  , src  , dstStride, srcStride);\
     OPNAME ## cavs_filt8_v_ ## NAME(dst+8, src+8, dstStride, srcStride);\
     src += 8*srcStride;\
@@ -320,7 +332,8 @@ static void OPNAME ## cavs_filt16_v_ ## NAME(uint8_t *dst, const uint8_t *src, i
     OPNAME ## cavs_filt8_v_ ## NAME(dst+8, src+8, dstStride, srcStride);\
 }\
 \
-static void OPNAME ## cavs_filt16_h_ ## NAME(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
+static void OPNAME ## cavs_filt16_h_ ## NAME(uint8_t *dst, const uint8_t *src, ptrdiff_t dstStride, ptrdiff_t srcStride)\
+{                                                                       \
     OPNAME ## cavs_filt8_h_ ## NAME(dst  , src  , dstStride, srcStride);\
     OPNAME ## cavs_filt8_h_ ## NAME(dst+8, src+8, dstStride, srcStride);\
     src += 8*srcStride;\
@@ -330,7 +343,8 @@ static void OPNAME ## cavs_filt16_h_ ## NAME(uint8_t *dst, const uint8_t *src, i
 }\
 
 #define CAVS_SUBPIX_HV(OPNAME, OP, NAME, AH, BH, CH, DH, EH, FH, AV, BV, CV, DV, EV, FV, FULL) \
-static void OPNAME ## cavs_filt8_hv_ ## NAME(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dstStride, int srcStride){\
+static void OPNAME ## cavs_filt8_hv_ ## NAME(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, ptrdiff_t dstStride, ptrdiff_t srcStride)\
+{                                                                       \
     int16_t temp[8*(8+5)];\
     int16_t *tmp = temp;\
     const int h=8;\
@@ -411,7 +425,8 @@ static void OPNAME ## cavs_filt8_hv_ ## NAME(uint8_t *dst, const uint8_t *src1,
     }\
 }\
 \
-static void OPNAME ## cavs_filt16_hv_ ## NAME(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dstStride, int srcStride){ \
+static void OPNAME ## cavs_filt16_hv_ ## NAME(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, ptrdiff_t dstStride, ptrdiff_t srcStride)\
+{                                                                       \
     OPNAME ## cavs_filt8_hv_ ## NAME(dst  , src1,   src2  , dstStride, srcStride); \
     OPNAME ## cavs_filt8_hv_ ## NAME(dst+8, src1+8, src2+8, dstStride, srcStride); \
     src1 += 8*srcStride;\
diff --git a/libavcodec/cavsdsp.h b/libavcodec/cavsdsp.h
index 847f5c4c87..9ccaa0addd 100644
--- a/libavcodec/cavsdsp.h
+++ b/libavcodec/cavsdsp.h
@@ -30,11 +30,11 @@
 typedef struct CAVSDSPContext {
     qpel_mc_func put_cavs_qpel_pixels_tab[2][16];
     qpel_mc_func avg_cavs_qpel_pixels_tab[2][16];
-    void (*cavs_filter_lv)(uint8_t *pix, int stride, int alpha, int beta, int tc, int bs1, int bs2);
-    void (*cavs_filter_lh)(uint8_t *pix, int stride, int alpha, int beta, int tc, int bs1, int bs2);
-    void (*cavs_filter_cv)(uint8_t *pix, int stride, int alpha, int beta, int tc, int bs1, int bs2);
-    void (*cavs_filter_ch)(uint8_t *pix, int stride, int alpha, int beta, int tc, int bs1, int bs2);
-    void (*cavs_idct8_add)(uint8_t *dst, int16_t *block, int stride);
+    void (*cavs_filter_lv)(uint8_t *pix, ptrdiff_t stride, int alpha, int beta, int tc, int bs1, int bs2);
+    void (*cavs_filter_lh)(uint8_t *pix, ptrdiff_t stride, int alpha, int beta, int tc, int bs1, int bs2);
+    void (*cavs_filter_cv)(uint8_t *pix, ptrdiff_t stride, int alpha, int beta, int tc, int bs1, int bs2);
+    void (*cavs_filter_ch)(uint8_t *pix, ptrdiff_t stride, int alpha, int beta, int tc, int bs1, int bs2);
+    void (*cavs_idct8_add)(uint8_t *dst, int16_t *block, ptrdiff_t stride);
     int idct_perm;
 } CAVSDSPContext;
 
diff --git a/libavcodec/x86/cavsdsp.c b/libavcodec/x86/cavsdsp.c
index e57d8be76a..ecb9b231f1 100644
--- a/libavcodec/x86/cavsdsp.c
+++ b/libavcodec/x86/cavsdsp.c
@@ -136,7 +136,7 @@ static inline void cavs_idct8_1d(int16_t *block, uint64_t bias)
     SBUTTERFLY(a,c,d,dq,q) /* a=aeim d=bfjn */\
     SBUTTERFLY(t,b,c,dq,q) /* t=cgko c=dhlp */
 
-static void cavs_idct8_add_mmx(uint8_t *dst, int16_t *block, int stride)
+static void cavs_idct8_add_mmx(uint8_t *dst, int16_t *block, ptrdiff_t stride)
 {
     int i;
     LOCAL_ALIGNED(16, int16_t, b2, [64]);
@@ -334,7 +334,8 @@ static void cavs_idct8_add_mmx(uint8_t *dst, int16_t *block, int stride)
    }
 
 #define QPEL_CAVS(OPNAME, OP, MMX)\
-static void OPNAME ## cavs_qpel8_h_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
+static void OPNAME ## cavs_qpel8_h_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t dstStride, ptrdiff_t srcStride)\
+{\
     int h=8;\
     __asm__ volatile(\
         "pxor %%mm7, %%mm7          \n\t"\
@@ -382,43 +383,53 @@ static void OPNAME ## cavs_qpel8_h_ ## MMX(uint8_t *dst, const uint8_t *src, int
     );\
 }\
 \
-static inline void OPNAME ## cavs_qpel8or16_v1_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride, int h){\
+static inline void OPNAME ## cavs_qpel8or16_v1_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t dstStride, ptrdiff_t srcStride, int h)\
+{                                                                       \
   QPEL_CAVSVNUM(QPEL_CAVSV1,OP,ff_pw_64,ff_pw_96,ff_pw_42)      \
 }\
 \
-static inline void OPNAME ## cavs_qpel8or16_v2_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride, int h){\
+static inline void OPNAME ## cavs_qpel8or16_v2_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t dstStride, ptrdiff_t srcStride, int h)\
+{                                                                       \
   QPEL_CAVSVNUM(QPEL_CAVSV2,OP,ff_pw_4,ff_pw_5,ff_pw_42)        \
 }\
 \
-static inline void OPNAME ## cavs_qpel8or16_v3_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride, int h){\
+static inline void OPNAME ## cavs_qpel8or16_v3_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t dstStride, ptrdiff_t srcStride, int h)\
+{                                                                       \
   QPEL_CAVSVNUM(QPEL_CAVSV3,OP,ff_pw_64,ff_pw_96,ff_pw_42)      \
 }\
 \
-static void OPNAME ## cavs_qpel8_v1_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
+static void OPNAME ## cavs_qpel8_v1_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t dstStride, ptrdiff_t srcStride)\
+{                                                                       \
     OPNAME ## cavs_qpel8or16_v1_ ## MMX(dst  , src  , dstStride, srcStride, 8);\
 }\
-static void OPNAME ## cavs_qpel16_v1_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
+static void OPNAME ## cavs_qpel16_v1_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t dstStride, ptrdiff_t srcStride)\
+{                                                                       \
     OPNAME ## cavs_qpel8or16_v1_ ## MMX(dst  , src  , dstStride, srcStride, 16);\
     OPNAME ## cavs_qpel8or16_v1_ ## MMX(dst+8, src+8, dstStride, srcStride, 16);\
 }\
 \
-static void OPNAME ## cavs_qpel8_v2_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
+static void OPNAME ## cavs_qpel8_v2_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t dstStride, ptrdiff_t srcStride)\
+{                                                                       \
     OPNAME ## cavs_qpel8or16_v2_ ## MMX(dst  , src  , dstStride, srcStride, 8);\
 }\
-static void OPNAME ## cavs_qpel16_v2_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
+static void OPNAME ## cavs_qpel16_v2_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t dstStride, ptrdiff_t srcStride)\
+{                                                                       \
     OPNAME ## cavs_qpel8or16_v2_ ## MMX(dst  , src  , dstStride, srcStride, 16);\
     OPNAME ## cavs_qpel8or16_v2_ ## MMX(dst+8, src+8, dstStride, srcStride, 16);\
 }\
 \
-static void OPNAME ## cavs_qpel8_v3_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
+static void OPNAME ## cavs_qpel8_v3_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t dstStride, ptrdiff_t srcStride)\
+{                                                                       \
     OPNAME ## cavs_qpel8or16_v3_ ## MMX(dst  , src  , dstStride, srcStride, 8);\
 }\
-static void OPNAME ## cavs_qpel16_v3_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
+static void OPNAME ## cavs_qpel16_v3_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t dstStride, ptrdiff_t srcStride)\
+{                                                                       \
     OPNAME ## cavs_qpel8or16_v3_ ## MMX(dst  , src  , dstStride, srcStride, 16);\
     OPNAME ## cavs_qpel8or16_v3_ ## MMX(dst+8, src+8, dstStride, srcStride, 16);\
 }\
 \
-static void OPNAME ## cavs_qpel16_h_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
+static void OPNAME ## cavs_qpel16_h_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t dstStride, ptrdiff_t srcStride)\
+{                                                                       \
     OPNAME ## cavs_qpel8_h_ ## MMX(dst  , src  , dstStride, srcStride);\
     OPNAME ## cavs_qpel8_h_ ## MMX(dst+8, src+8, dstStride, srcStride);\
     src += 8*srcStride;\